4,624,557 AI resources indexedWorking toward the world’s largest AI resource index.

work-unit-decomposition

Decompose implementation plans into discrete work units with enumerated DoD items, file scope declarations, dependency mapping, and human checkpoint flags.

Installable skillMITBy a5c-aiUpdated Oct 28, 2026
INTENDED FOROpenClawInstall in a listed agent.

instructions.md

---
name: work-unit-decomposition-cb86766ca10c
description: Decompose implementation plans into discrete work units with enumerated DoD items, file scope declarations, dependency mapping, and human checkpoint flags.
allowed-tools: Read, Write, Edit, Bash, Grep, Glob, WebFetch, WebSearch, Agent, AskUserQuestion
graph:
  domains: [domain:software-engineering]
  skillAreas: [skill-area:agentic-loops, skill-area:orchestration-loop]
  workflows: [workflow:feature-development]
  topics: [topic:developer-experience]
  roles: [role:tech-lead, role:backend-engineer]
---

- When decomposing a large feature into implementable chunks
- When preparing for orchestrated execution

## Work Unit Structure

Each work unit specifies:
- **ID** - Unique identifier
- **Title** - Human-readable description
- **Definition of Done** - Enumerated checklist items
- **File Scope** - Which files the unit may modify
- **Dependencies** - Other work units this depends on
- **Human Checkpoint** - Whether human approval is needed before execution
- **Parallel Safe** - Whether it can execute alongside other units

## Human Checkpoint Triggers

Mark human checkpoints for:
- Schema or database changes
- Security-sensitive code paths
- New architectural patterns not seen in codebase
- External API integrations
- Breaking changes to public interfaces

## Agents Used

- `agents/architect/` - Creates the decomposition
- `agents/cto/` - Validates TDD readiness per unit

## Tool Use

Invoke as part of: `methodologies/metaswarm/metaswarm-orchestrator` (Phase 4)

Platform guidance

Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.